Regression to the mean applies any time a process has high variance. If I have an unusually productive week, odds are the next week will be worse, simply because the previous one was an outlier. But people often make the mistake of noticing week X+1 was much worse than week X, and attributing it to some variable that changed between X and X+1.